ofermend

Given that coding agents like Claude Code or Codex are now quite good, there's of course a massive increase in PRs submitted to open source projects. Not all of them are great, some are true AI slop.

See for example Daniel Stenberg on the topic as it relates to cURL: https://daniel.haxx.se/blog/2026/01/26/the-end-of-the-curl-bug-bounty/ and https://daniel.haxx.se/blog/2025/07/14/death-by-a-thousand-slops/

I'm curious how much of a problem this is to other open source projects

Committers, how much of this pain are you seeing, and are you using any AI tools to mitigate or address this issue?

We often debate "can humans trust AI?" An equally important question: can AI trust humans?

Shadow AI = Shadow IT 2.0

Large companies implementing generative AI are experiencing the re-emergence of all the issues and headaches we know well and remember, that were associated with "shadow IT".

This time, applied to RAG, LLMs and agents.

Curious how people are addressing this?